Benchmark Radar

Every value that could be read verbatim from a cited document, placed at the date that document was published rather than at any evaluation date. The line directly links actual observations that set a new reported record among the values shown; it neither holds a score between reports nor extends past the final record. Test versions and run conditions can differ, so this is a reported-record path, not a like-for-like trend. When no newer number could be read, the gap is marked rather than drawn through. Whether a benchmark has saturated stays a reading you make, not a score this panel prints.

Two scores can only be compared when each one records which version of the test it used, which slice of it, whether a high number is good or bad, which model, what software ran it, how much thinking time it was given, what it cost, how long it took, when it was published, and where the number came from. The test and the way it was run have to match; the model, cost, time and date are free to differ, because those are what the chart compares. This site records that a model card mentioned a benchmark. It does not yet record those measurements. Once they exist, this chart could plot cost or speed against score, and draw a line through only the results that nothing else beats on both at once. A date slider would then show how that line moved over time.

Built for AI researchers and model evaluators: filter LLM, agent, multimodal and AI-for-science benchmarks by date, source and score to compare evals quickly.

Each benchmark has its own page listing the paper, code, dataset and reported scores. A models grid puts models in columns and benchmarks in rows, so you can select any score to see its setting and source. The saturation views chart reported scores over time and rank benchmarks by a 30-day attention score built from GitHub stars and Hugging Face signals.

Benchmark Radar is designed for a range of users, including: • AI researchers choosing evaluation benchmarks for new models • ML engineers comparing reported model scores across benchmarks

With Benchmark Radar you can: • Finding which benchmarks have reported scores for a given model or task area. • Tracking newly released AI benchmarks and their attention over the last 30 days. • Checking whether a benchmark is nearing saturation by reviewing its reported score history.